genkit
Pass
Audited by Gen Agent Trust Hub on Mar 11, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill fetches the Genkit CLI installation script from the official domain
cli.genkit.devvia a piped shell command. - [COMMAND_EXECUTION]: Provides standard instructions for installing and running Node.js development tools, including
genkit-cli,tsx, and various Genkit model/vectorstore plugins. - [PROMPT_INJECTION]: The skill documents the creation of AI flows that ingest and process external user-provided data.
- Ingestion points: Found in
summarizeFlow,agentFlow, andragFlowwhere untrusted text is interpolated into LLM prompts. - Boundary markers: Present in
ragFlowandchatFlowexamples using system instructions to ground model responses in context. - Capability inventory: Utilizes
ai.generatefor text generation andai.defineToolfor executing functional logic. - Sanitization: Uses Zod schemas for strict input/output structure validation, although natural language content sanitization is the responsibility of the implementer.
Audit Metadata